Approximately 22 million Americans have a hearing impairment. While hearing devices such as hearing aids (HAs) and cochlear implants (CIs) are successful in improving speech recognition for many hearing-impaired (HI) individuals, there is still significant variability in benefit, and speech recognition in noise remains a problem. One factor that may limit benefit is interference between bilaterally worn hearing devices, whether the devices are bilateral HAs, bilateral CIs, or a unilateral CI worn with a HA in the contralatera ear (bimodal CI+HA). In HI individuals, spectral information is not necessarily matched between ears due to asymmetries introduced by hearing loss and hearing device programming. Thus, unlike normal-hearing (NH) listeners, HI listeners integrate mismatched spectral information between ears, in a process that we can call binaural spectral integration (BSI). Our preliminary studies suggest that BSI adapts to reduce perception of interaural spectral mismatch in HI individuals; binaural fusion is increased so that pitches that differ greatly in frequency between the two ears are still heard as a single percept. However, a side effect of increased binaural fusion is that mismatched spectral information is averaged between ears, which may lead to increased speech perception confusions and worsened frequency discrimination. The long-term goal of this research program is to investigate how BSI differs between HI and NH listeners, and the clinical implications of these differences for rehabilitation. Specifically, in this proposal w will examine 1) differences in binaural fusion between HI and NH listeners, as well as across different device combinations within the HI group: bilateral HAs, bilateral CIs, and bimodal CI+HA; 2) how abnormal BSI affects the perception of tone and speech stimuli; and 3) how BSI changes as children develop, because developing brains may adapt BSI differently than adults. The proposed research will increase our understanding of how one aspect of auditory processing, binaural spectral integration, contributes to difficulties with binaural speech perception in HA and CI users. The results will also indicate whether hearing devices should be programmed differently for children than adults. Finally, the results will build the theoretical framework for future planned studies using frequency-specific device programming and training programs to reduce any detrimental BSI, while maintaining or increasing the benefits of binaural processing for speech perception in quiet and noise.